Improving and tightening my overclock...

Yeggstry

Yeggstry

Yeggstry

Associate
Joined
8 Aug 2005
Posts
1,101
Location
Reading, UK
Hi guys,

I have some questions with regards to improving my clock up to 3.6GHz - realistically the highest amount I can safely get the Q6600 G0 up to (rest of the spec in my sig).

I currently have the following settings in the BIOS:

FSB: 377
Multiplier: 9
Vcore: 1.40v
DRAM Volt: 2.10v
RAM Timings: 5-5-5-15

The rest is set to auto. The following is reported on windows at 100% processing - that's folding, not prime (it is prime stable though):

VID: 1.2375v

Temps
Cores: 63 62 54 52 (note: these temps are about 8-10C less than when testing for prime stability)
NB: 43
SB: 37

Voltages
Vcore: 1.32v
SB: 1.09v
FSBT: 1.42v
DRAM: 2.19v
DRAM Ref: 1.10v
NB: 1.49v


I plan on putting my new hard drive for RAID 0 shortly, and I'm going to take a look at my spread of my thermal paste as I think it may not be covering one side of the processor, doubt it will make much difference though.

My question is, what should I change in order to get myself stable to 3.6? I tried previously with 400 * 9 at 1.425v and it lastest for a couple of hours through prime, but crashed shortly after that. I also noticed that the NB voltage was being reported as 1.69v at the time - not good! The DRAM voltage seems a bit high at the moment as well, 2.1v is the max rated voltage...

Is it also worth me trying 450 * 8 rather than 400 * 9? What would the difference be voltages and temp wise? I'm thinking about this because the RAM I have can be set up to 1000 in the BIOS without overclocking, and would the machine be quicker with higher FSB, lower multiplier and a higher RAM clock or with a lower FSB, higher multiplier and lower RAM clock but tighter timings i.e. 4-4-4-12?

Any help would be welcome :)

p.s. this isn't a reactionary thread from Cob's warning about his quad core, I've been planning this for a while :p
 
i would suggest the difference in temps is due to the concave shape of the top of your cpu and also the same on any heatsink (which one do you have??). Consider lapping your cpu/heatsink to reduce temps (i.e better thermal conductivity) which should also reduce the difference between the cores.

Gerenally your temps are ok at the moment - max i would say long term would be 65 degrees (also with the summer etc temps are likely to rise a fair bit)

If you can tighten the timings to 4-4-4-12 then that would help a touch. I think there is not a huge difference between the higher clocks/tighter timings issues - did a small 12 hour test myself and noticed negligable change in ppd with higher clock/slack timings and lower clock/tighter timings. I think clock speed is always the best option.

Also sometimes that extra 200mhz or so sometimes needs a huge jump in voltages etc to get there.

I know i have a fairly old Q6600 as the VID is 1.325v and i need 1.425v for 3.15ghz stable (this is my 24/7 folding speed) but i need 2.2V on the ram and 1.525v for 3.33ghz which takes the temps to near 70 degrees so just not worth it for another 180mhz or so.

Good luck getting there but don't be to hard on yourself for another 200mhz!!
 
Ok, I tried the following a couple of nights ago and it BSODed on me fairly quickly:

FSB 400
RAM 5-5-5-15
Vcore 1.45v
DRAM 2.10v
NB 1.45v
FSBT 1.24v, then tried Auto

Is it simply a case of me upping the Vcore? I've had a look around and seen that guys with similar VIDs have been getting 3.6 on about 1.45v.
 
We seem to have a slacking news reporter :P

I've upped the voltage to 1.475v in BIOS, which is reporting as 1.384v in CPUZ. Been folding for 90 mins so far, and the temps aren't that much worse :) 65,63,62,61. If this stays stable overnight I'll prime stable it then see how much more I can push it.
 
You must log in or register to reply here.
Back
Top Bottom